Tencent has signed a contract worth around $7bn to access 100,000 GPUs through Oracle's cloud platform, according to a Financial Times report citing two people familiar with the matter.

The deal, a five-year lease across multiple Oracle data centers in Southeast Asia, gives Tencent access to GPU capacity it cannot obtain in China because US export controls restrict the hardware. Tencent is paying 30 percent of the roughly $7bn value upfront.

Tencent intends to use the compute to advance its AI models and agentic tools. DCD has contacted Oracle and Tencent for comment.

Why is Tencent spending this much now?

The company's most recent quarterly results show the scale of its infrastructure commitments. Tencent reported negative free cash flow of RMB 13.8 billion ($2.06bn) for its second quarter — the company's first negative figure in more than a decade, according to the FT.

CFO John Lo said at the earnings call that the result was "reflecting large AI infrastructure capex and AI-related prepayments." He added that "excluding the prepayments for compute procurement, free cash flow would have been RMB 37.6bn ($5.61bn)," a difference of roughly $7-8bn — in the same range as the reported Oracle deal.

Tencent president Martin Lau framed the spending as a low-risk bet: "We're comfortable in making significant investments in AI because not only is there a substantial upside potential, there is also clear downside protection."

"The AI investments we're making are mostly in AI infrastructure, and in the worst case, which we do not believe would happen, we can choose to rent that infrastructure out at cost recovery or even better prices via Tencent Cloud if needed," Lau said.

What will the compute be used for?

Executives outlined a clear sequence for the new capacity:

• First: "training bigger and better Hunyuan models"
• Then: inferencing workloads
• Later: renting out capacity as bare metal or via Model-as-a-Service offerings

This matches Tencent's previously stated strategy of prioritizing internal services over external cloud monetization.

CSO James Mitchell said in May 2026: "Looking through the rest of this year, as the supply of China-designed GPUs progressively ramps up, we'll be remedying that situation; we will be making more capacity available in Tencent Cloud, and consequently driving up Tencent Cloud's rate of expansion."

"That's where the trade-off has been made: we have been consciously late to monetize the AI opportunity through Tencent Cloud because we've been simultaneously supporting a number of AI initiatives internally," Mitchell said.

The Oracle lease addresses the immediate supply gap: US export controls prevent Tencent from buying the GPUs outright in China, but cloud access from data centers in Southeast Asia offers a workaround at scale. The 30 percent upfront payment ties a substantial portion of Tencent's reported prepayment outflow directly to compute procurement.

For Oracle, the contract adds a multi-billion-dollar, multi-year commitment in a region where it operates several data center sites. For Tencent, it secures GPU capacity sized in the six figures at a moment when its domestic supply of China-designed GPUs is still ramping up.

The arrangement also signals that aggressive AI infrastructure leasing — rather than chip ownership — remains a viable path for Chinese hyperscalers operating under US export restrictions, at least while domestic GPU supply matures.